.0001" marker marks

DMF_TomB · Dec 17, 2017

anybody know of a marker where if you make a mark on a part it adds .0001" thickness to it. that is instead of shimming between parts just making a mark where the ink thickness is .0001"
.
had a 1 ton part and sitting on 4 supports and probe says one support is .00045" high (.0003" tolerance). i made a sharpie marker mark but didnt seem to do much.
.
i ended up on high support (6 square inches) marking up with a marker and sanding mark off and it took 3 times to get .00045 down to .0002" (100 grit sandpaper)
.
i have heard of grinders using a crayon to add thickness instead of shims. just seems a wax crayon would leave a uneven thickness.
.
using 2 ea .03mm (.0012") shims on one end of part (.0024") and .001 and .0015" on other end (.0025") gives .0001" difference but sometimes you want to use no shims

Finegrain · Dec 17, 2017

Found this: How Thick is the Line from a Sharpie? - Core77

DMF_TomB · Dec 17, 2017

i made red sharpie marks and probe says it made less than 50 millionth difference with 1 ton part. perhaps under weight it flattens out a bit.
.
if i mark up surface (6 square inches) and sand off mark with 100 grit sandpaper it seems to have about 50 millionths effect or .00005" or do 3 times removes .00015" of the steel part

706jim · Dec 17, 2017

The video explains the line thickness very well. However, my concern would be the effect of clamping compression on the ink line unless the ink covered the complete surface in contact. If it didn't, then I'd think the ink layer would compress to less than his calculated value. (Which for those of you who don't care to watch the video range from just over one tenth for Sharpies and approximately double that for Dykem.)

Richard King · Dec 17, 2017

I have used Dykem spray bluing to get a tenth on test bars I need to use to tram Horizontal spindles on Mills.

Monarchist · Dec 17, 2017

Richard King said:
I have used Dykem spray bluing to get a tenth on test bars I need to use to tram Horizontal spindles on Mills.

+1

I'd guess the key there is that layout coatings - exact opposite of high-spot coatings - are formulated to pull themselves into even-thicknesses as much as they can do, then attempt to be enduringly stable in that state until intentionally removed.

DMF_TomB · Dec 17, 2017

perhaps when about 30 lbs per sq in pressure put on it the coating flattens out a bit. when i put a 1 ton part on 4 supports each about 15 square inches part did not probe as higher on that support with probing repeating .00005"
.
i used red sharpie to markup highest support and sanded off with 100 grit and part probed .0001" lower and 2 more markups and sanding part probed .00015" lower. initially it was .00045 high and i got it down to .0002" on the highest support (3 sandings of red sharpie markups with 100 grit)
.
i made 2 layers of red sharpie marker mark on lowest support and it had less than .00005" shim effect perhaps cause of the weight involved.

Colt45 · Dec 17, 2017

My experience has been that a sharpie is typically 50 millionths and that it can vary depending on the material and surface finish. Have had trouble getting "brush on" dykem at any kind of uniform thickness I felt was accurate. and repeatable. There is a guy on Ebay who sells shims in tenths increments, might want to look him up.

cameraman · Dec 17, 2017

Oxtoolco sharpie video

I Remember OxTool.Co Did a youtube video on Sharpie thickness, with additional "Extra scientific" statistics :-) + thickness of different colors.
AND blue and red Dykem (as quick compare).

How thick is a sharpie mark? - YouTube

^^Kinda fun but definitely makes you think.

hanermo · Dec 17, 2017

Mostly as cameraman said.

Area/thickness of line or lines, have a pwr 2 relationship in area vs pressure exerted, scaling at pwr 3 or pwr 4.
So its complex.
And as the youtube video shows (many others as well) markers have widely varying thicknesses.

This is how I see it.
Any really significant marking via sharpie, dykem, or shims will always be dependant on the load, and the area of shim vs total surface vs force applied.
So 2 lines for the same surface will resist 2x the force.

For 50+ years we have known shims work pretty well.
They compress fairly easily upto a certain point, and then a tiny bit, and then almost nothing until huge loads are applied.
The "next yield point" for shims be it sharpie or cigarette paper is similar to coining or hydroforming vs a press pushing down on something.
This load to achieve coining or liquid transformation is typically 10-50x more than a major press operation ... from what I have seen.
It took 5000 m / = 500 bars to transform a metal watch case for example (held it in my hands).

The OP wanted 0.0001" or 2 micron marks.
I feel pretty confident that needs an industrial process of some type, vapour disposition, vacuum, electron, etc.
Shims to 0.01 mm are cheap, but better ?

Are both surfaces flat enough ? Sure ? Why ?
If a corner is carrying ... how will You know ?

jrmach · Dec 17, 2017

Cigarette pack cellophane is a good cheap thin shim

Monarchist · Dec 17, 2017

jrmach said:
Cigarette pack cellophane is a good cheap thin shim

Cheap? Only if one doesn't have to buy the pack of cigarettes just to get it.

Thin? Only relatively so.

Good? Not so much at all.

OP is (attempting / discussing / conspiring at) working well below the ranges we've "solved" with various papers, foils, foil-coated papers, and plastic films for Donkey's Years.

There is very little consistency, predictability, or repeatability to most of the "shade tree" favourites. Folks just "made do" better than nothing-at-all wudda done. Or convinced ourselves we had.

machtool · Dec 18, 2017

DMF_TomB said:
perhaps when about 30 lbs per sq in pressure put on it the coating flattens out a bit. when i put a 1 ton part on 4 supports each about 15 square inches part did not probe as higher on that support with probing repeating .00005"

Aren't you still on those 25 year old Toshiba's? Who's resolution only goes down to 4 decimal places in Imperial. Yet your quoting .00005" 5 decimals. Half a tenth. Its the Dunning - Kruger syndrome kicking in yet again. You cant possibly see 5 decimals on your control? Tell us how your probe repeats until .00005"

All this in an un-stabilised / un-air conditioned machine hall. Your trying to tell us, that you can lift this 1 ton shit up, scribble a few Sharpie lines and can move location references by less than half a tenth.

All the time, grinding / lapping / stoning with a bit of 100 grit sandpaper. Your not fooling anybody. Truth by known, there is no way in hell you could un-clamp, crane, re-clamp that casting with in two thou, little own half a tenth. Its a 25 year old gantry mill for fucks sake, and its not climate controlled. Any one that claims half a tenth is a dickhead. I don't care what the control says. Even thou the control only goes to tenths. (X31.0000)

I get the feeling, every time you come here with one of these dumb questions, your only looking for justification as to why to not get sacked.

These 4 supports your mention that are .00045 (4 and a half tenths), out of kilter, surely Gleason machines the reference pads in house. I'd predict they are linear bearing block mounts), They didn't come in that way from the raw casting plant in Taiwan, I'd predict his name is Tony. Wouldn't these 4 reference pads, which you say are out, be machined on one of your other qualifying machines? You cant turn raw castings into shit with out pre-machining.

Why don't you talk to the bloke, that machines those reference pads. Tell him his pads are 4 & a HALF TENTHS OUT. That will save you scribbling on them with a sharper marker. Given the size of your mountainous casting's, that you like to tell us about. That should sort it. If he could get them 0.0000000 you'd be hard pressed to have to shim them, or scribble on them?

Your milking that job. Large machine, no climate control, inventing decimal place's the control doesn't have, and you want to talk about spitting the atom.

I'd sack you on the spot.

machtool · Dec 18, 2017

machtool said:
Aren't you still on those 25 year old Toshiba's? Who's resolution only goes down to 4 decimal places in Imperial. Yet your quoting .00005" 5 decimals. Half a tenth. Its the Dunning - Kruger syndrome kicking in yet again. You cant possibly see 5 decimals on your control? Tell us how your probe repeats until .00005"

All this in an un-stabilised / un-air conditioned machine hall. Your trying to tell us, that you can lift this 1 ton shit up, scribble a few Sharpie lines and can move location references by less than half a tenth.

All the time, grinding / lapping / stoning with a bit of 100 grit sandpaper. Your not fooling anybody. Truth by known, there is no way in hell you could un-clamp, crane, re-clamp that casting with in two thou, little own half a tenth. Its a 25 year old gantry mill for fucks sake, and its not climate controlled. Any one that claims half a tenth is a dickhead. I don't care what the control says. Even thou the control only goes to tenths. (X31.0000)

I get the feeling, every time you come here with one of these dumb questions, your only looking for justification as to why to not get sacked.

These 4 supports your mention that are .00045 (4 and a half tenths), out of kilter, surely Gleason machines the reference pads in house. I'd predict they are linear bearing block mounts), They didn't come in that way from the raw casting plant in Taiwan, I'd predict his name is Tony. Wouldn't these 4 reference pads, which you say are out, be machined on one of your other qualifying machines? You cant turn raw castings into shit with out pre-machining. Have you made it aware to management that your datum's are all out of kilter?

Why don't you talk to the bloke, that machines those reference pads. Tell him his pads are 4 & a HALF TENTHS OUT. That will save you scribbling on them with a sharper marker. Given the size of your mountainous casting's, that you like to tell us about. That should sort it. If he could get them 0.0000000 you'd be hard pressed to have to shim them, or scribble on them?

Your milking that job. Large machine, no climate control, inventing decimal place's the control doesn't have, and you want to talk about spitting the atom.

I'd sack you on the spot. Your seeing error's on a big machine, which your feeble brain can't compute.

Run us past it again. Who produces these reference surfaces? The ones we chew with 100 grit, or slather with Sharpie Marker. Post # 3 you were claiming 50 millionths.

DMF_TomB said:
i made red sharpie marks and probe says it made less than 50 millionth difference with 1 ton part. perhaps under weight it flattens out a bit.

Can some one explain digital rounding to this cat, He doesn't have that resolution. By a long shot.

DMF_TomB · Dec 18, 2017

1) trim pads are fixture pads (about 2x3") part sits of that get .0010" machined every time fixture is put on the table and probe confirms flat within .0003"
.
2) i remove previous 1 ton part that was in fixture. i blow off and clean fixture pads and clean bottom of next 1 ton part and set on fixture pads and probe probes bottom of part. reading can vary .0005" between probing from thermal changes in machine but the readings from one higher pad to the next lower pad relationship stays same. if one pad is .00045" higher it repeats .00005" or 50 millions. probe takes 2 readings 0.1" apart and averages readings as part surface is wavy as much as .0003". thats why 2 readings taken at each corner of 40" part to help average readings cause part wavy. i write probe reading down as it tells me exactly which trim pad is high and by how much. i repeat probe reading before removing part, relationship of highest to lowest pad repeats to .00005"
.
3)i marked up with red sharpie the low pad and set part on pads and probe and it had .00005" or less effect comparing low pad to high pad probe reading relationship repeats about .00005", i did probing 2 times confirming repeatability
.
4) i red sharpie high pad and using 100 grit remove totaly red sharpie mark and set part down and reprobe. relationship of high pad to lower pad went .00045 to .00035, i repeat probe readings and get .000325", i removed part and marked up and sanded 2 more times and set part down and reprobed and it went .00035 to .0002 high to lowest pads
.
5) of course i could have just did the trim op and machined milled .001 off each pad. sometimes that works and sometimes not. machine usually leave one particular corner trim pad .0002 to .00025 higher, part of reason we have a 5 minute spindle running warmup is as spindle warms up the last pad machined is lower than first pad done cause spindle getting longer as it warms up
.
6) dont care if you dont believe what i say. i just asked if a marker could apply .0001" by the thickness of the marks applied. for me with 30 to 100 psi pressure it appears the mark flows a bit into the milled surface and it will not reliably apply .0001" like a shim
.
7) i probe bottom of 40" square part (about 16" deep) sitting on 4 pads. most likely the machine parts bottom surface that gets probes is wavy usually less that .0003" but added to the .0003" tolerance of the 4 trim pads machined it can combine. reason i never see .0006 error is program probes different spots and averages readings. thus why i sometimes see .0004 or .00045" error
......i tried the sharpie mark as a shim and i tried sanding sharpie mark off to remove .0001, is it faster than remachining the 4 trim pads. seems about the same. i took extra time rerunning the probing checking for repeatability. took time picking part up 2nd time cleaning again and setting down again and reprobing 2 more times confirming parts cleaned. if i went directly to sanding sharpie mark off 3 times i probably would have been done 6x faster. i did not known how much the sanding red sharpie mark off would remove. with 100 grit it removes .00005 to .0001" each time.......when one pad is .0001" high it can be time consuming. let just say i cleaned the highest pad and made it extra extra clean with the 100 grit sand paper (lapping block). we used to call it using hard solvent to clean

Monarchist · Dec 18, 2017

DMF_TomB said:
1) trim pads are fixture pads (about 2x3") part sits of that get .0010" machined every time fixture is put on the table and probe confirms flat within .0003"
.
2) i remove previous 1 ton part that was in fixture. i blow off and clean fixture pads and clean bottom of next 1 ton part and set on fixture pads and probe probes bottom of part. reading can vary .0005" between probing from thermal changes in machine but the readings from one higher pad to the next lower pad relationship stays same. if one pad is .00045" higher it repeats .00005" or 50 millions. probe takes 2 readings 0.1" apart and averages readings as part surface is wavy as much as .0003". thats why 2 readings taken at each corner of 40" part to help average readings cause part wavy. i write probe reading down as it tells me exactly which trim pad is high and by how much. i repeat probe reading before removing part, relationship of highest to lowest pad repeats to .00005"
.
3)i marked up with red sharpie the low pad and set part on pads and probe and it had .00005" or less effect comparing low pad to high pad probe reading relationship repeats about .00005", i did probing 2 times confirming repeatability
.
4) i red sharpie high pad and using 100 grit remove totaly red sharpie mark and set part down and reprobe. relationship of high pad to lower pad went .00045 to .00035, i repeat probe readings and get .000325", i removed part and marked up and sanded 2 more times and set part down and reprobed and it went .00035 to .0002 high to lowest pads
.
5) of course i could have just did the trim op and machined milled .001 off each pad. sometimes that works and sometimes not. machine usually leave one particular corner trim pad .0002 to .00025 higher, part of reason we have a 5 minute spindle running warmup is as spindle warms up the last pad machined is lower than first pad done cause spindle getting longer as it warms up
.
6) dont care if you dont believe what i say. i just asked if a marker could apply .0001" by the thickness of the marks applied. for me with 30 to 100 psi pressure it appears the mark flows a bit into the milled surface and it will not reliably apply .0001" like a shim
.
7) i probe bottom of 40" square part (about 16" deep) sitting on 4 pads. most likely the machine parts bottom surface that gets probes is wavy usually less that .0003" but added to the .0003" tolerance of the 4 trim pads machined it can combine. reason i never see .0006 error is program probes different spots and averages readings. thus why i sometimes see .0004 or .00045" error
......i tried the sharpie mark as a shim and i tried sanding sharpie mark off to remove .0001, is it faster than remachining the 4 trim pads. seems about the same. i took extra time rerunning the probing checking for repeatability. took time picking part up 2nd time cleaning again and setting down again and reprobing 2 more times confirming parts cleaned. if i went directly to sanding sharpie mark off 3 times i probably would have been done 6x faster. i did not known how much the sanding red sharpie mark off would remove. with 100 grit it removes .00005 to .0001" each time.......when one pad is .0001" high it can be time consuming. let just say i cleaned the highest pad and made it extra extra clean with the 100 grit sand paper (lapping block). we used to call it using hard solvent to clean

Nicely "logged" and explained monkey-motion, but still.... 100 grit abrasive is NOT joined-at-the-hip to Sharpie marks (alone), "wishing don't make it so", nor pompous rationalization - ergo it isn't really anything else BUT monkey-motion.

Least of all "scientific", nor a reliable guide to others any more than a Machinist of an earlier generation wiping a smear of sebum, dead skin flakes, bacteria, and coal-fire soot off the side of his nose onto the pad and wiggling the workpiece a few times wudda been...

IOW .. yer playin' wit yerself. And... it has gone more boring than amusing.

David_M · Dec 18, 2017

Fine diamond lapping paste (if the company is buying. Actually, is isn't that costly in the small amount needed). Coarseness determines thickness

Edit:

1800 grit is equal to .0001". Use the paste as the shim. It will spread out evenly and the diamond grit won't compress.

DMF_TomB · Dec 18, 2017

red sharpie mark like any plastic tends to flow under pressure. thus why 500 lbs on one pad flattened the sharpie mark for me down to .00005" or less. i have seen before polyurethane repairs to steel parts that sometimes work for decades and i have also seen where under a heavy load the repair breaks down in a few minutes. compression strength is only so high for a ink from a sharpie marker
.
sanding a red sharpie mark off to remove .00005" to .0001" of metal whether you call it lapping, hard cleaning, etc has been around long before sharpie markers were 1st made.
.
but i figured i would try both methods after all the probe readings of the part bottom sitting on the fixture pads will verify it. the relationship from high pad to lower pad repeats usually to .00005". i have done easily over 500 parts in the tonnage range. i got documentation on removing a part and next part bottom reads exceeding .0003" probe tolerance about 30% of the time (1st part bottom within .0003 and next part reading .0004).
.
rerunning the .001 trim milling pass does not always fix the problem. literally 100% of the time, the same corner trim pad reads higher than other pads. only question is does it read .0002 or .00045 higher ? like i said most likely part bottom being probed is wavy. readings taken 0.1" apart often vary .0001 to .0002" and the 2 readings are averaged (literally probe is measuring waviness). of the 2 reading taken 0.1" apart there is a repeatable higher and lower one, relationship repeats .0001 or closer
.
my question was only is there a marker mark (brand of marker) than reliably makes .0001" thick marks. for me cause of the part weight a red sharpie marker mark is not reliable enough for a .0001" shim.
.
i dont care much of opinions of how i do my job. only the inspection people who check my parts after i am done is what i care about. they will give me a copy of the full inspection report and i will ask if needed a explanation of exactly what it means. we have machine error compensation required in setup instructions when needed. basically sometimes there is a +/- tolerance and we need to favor a particular direction. normally we only compensate .0001 to .0003 per 40"
.
i will be running a new $2,400,000. cnc mill in a few months and it will be interesting to see if it holds tolerances any better than the other machines i run. i suspect part warpage and surface waviness will not go away but i will record what it does

DMF_TomB · Dec 18, 2017

David_M said:
Fine diamond lapping paste (if the company is buying. Actually, is isn't that costly in the small amount needed). Coarseness determines thickness

i use 3M Stikit paper on a precision ground block of steel for a lap.
.
basically Stikit paper is a adhesive backed roll of sandpaper. it comes in different grits. 100 grit is common for removing mill marks. after 10 to 60 seconds the milled surface marks are gone depends on how rough the surface was.
.
i have also used 400 grit when i had equipment that needed less than .00005" removed. i used a computer controlled capacitance gage to measure slot widths on paper coating equipment to .00001" tolerances. we would print out a 60" long print out next to 60" long part and it would show magnified where it needed sanding so slots were more parallel. why slot go out of parallel ? over time probably from abrasive chemical flow through slots causing more wear in spots. the coater would take 7 individual layers of chemical parallel pouring on to paper moving at 1100 feet per minute to apply the chemical layers to photographic paper. technology been around easily over 60 years
.
sandpaper on flat block is also to aid sanding and also used on pouring lip which needs a precise angle so not get a disturbance in the 7 layers of chemical pouring parallel like a water fall about 10", the pouring lip when cleaned often the operators damaged it and it needed lapping to repair it

dcsipo · Dec 18, 2017

DMF_TomB said:
anybody know of a marker where if you make a mark on a part it adds .0001" thickness to it. that is instead of shimming between parts just making a mark where the ink thickness is .0001"
.
had a 1 ton part and sitting on 4 supports and probe says one support is .00045" high (.0003" tolerance). i made a sharpie marker mark but didnt seem to do much.
.
i ended up on high support (6 square inches) marking up with a marker and sanding mark off and it took 3 times to get .00045 down to .0002" (100 grit sandpaper)
.
i have heard of grinders using a crayon to add thickness instead of shims. just seems a wax crayon would leave a uneven thickness.
.
using 2 ea .03mm (.0012") shims on one end of part (.0024") and .001 and .0015" on other end (.0025") gives .0001" difference but sometimes you want to use no shims

Why don't you mount .250" thick aluminum pads on the supports and lob them off with a face mill before placing the part on the supports? That way you know that they will be in the same plane within the accuracy of your mill. You should be able to verify them with the probe. If the measurements are off than you know that is the tolerance of your mill. If it cannot keep the tolerance machining aluminum all betts are off on CI or steel.

dee
;-D

.0001" marker marks

Diamond

Diamond

Diamond

Titanium

Diamond

Diamond

Diamond

Hot Rolled

Diamond

Titanium

Titanium

Diamond

Diamond

Diamond

Diamond

Diamond

Hot Rolled

Diamond

Diamond

Diamond

Similar threads